Triple
T15338099
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Cheltenham Gold Cup |
E366719
|
entity |
| Predicate | notableWinner |
P2766
|
FINISHED |
| Object | Kauto Star |
E364773
|
NE FINISHED |
How this triple was built (2 steps)
Every LLM step that produced this triple, in pipeline order — named-entity classification, the disambiguation choices (the exact options shown, with the pick highlighted), and the generated description. The batch + timestamp of each is in the Provenance table below.
NER
Named-entity recognition
gpt-5-mini
Instruction
Given a phrase, classify it is english named entity (e.g., persons, organizations, works of art) in Latin script, or not (e.g., literals, dates, URLs, verbose phrases). For disambiguation, the statement where the phrase occurs as object is also given. Please return a JSON object with `phrase` (string, the phrase being analyzed) and `is_ne` (boolean, indicating whether the phrase is a Named Entity).
Input
Phrase: Kauto Star | Statement: [Cheltenham Gold Cup, notableWinner, Kauto Star]
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Kauto Star Context triple: [Cheltenham Gold Cup, notableWinner, Kauto Star]
-
A.
Kauto Star
chosen
Kauto Star was a legendary French-bred, British-trained National Hunt racehorse best known for winning multiple King George VI Chases and Cheltenham Gold Cups.
-
B.
Katran
Katran is a high-performance, open-source layer 4 load balancer developed by Meta that leverages eBPF for efficient packet processing in large-scale network infrastructures.
-
C.
Rockhal
Rockhal is a major contemporary music venue and cultural center in Luxembourg, known for hosting international concerts and events.
-
D.
Stryama
Stryama is a river in Bulgaria that flows through the central part of the country before joining the Maritsa River.
-
E.
Karavola
Karavola is the summit that forms the highest peak of Mount Parnitha in Greece.
- F. None of above.
- G. Unsure - the case is ambiguous/there is not enough information to decide.
Provenance (3 batches)
The batch behind each pipeline step, in order, with when it ran. Timestamps are batch-level — stages were processed in waves, so the object chain (NER → NED1 → NEDg → NED2) reads in order, but predicate / elicitation batches can sit in a different wave.
| Step | Stage | Batch ID | Status | When |
|---|---|---|---|---|
| creating | Elicitation | batch_69d85a1355608190a6673ddb67231d54 |
completed | April 10, 2026, 2:01 a.m. |
| NER | Named-entity recognition | batch_69e03e11b22c81908280efe65acd5454 |
completed | April 16, 2026, 1:40 a.m. |
| NED1 | Entity disambiguation (via context triple) | batch_69ff01f2ee9c819080fce24ed13a07c7 |
completed | May 9, 2026, 9:44 a.m. |
Created at: April 10, 2026, 3:17 a.m.